Block space needs to be limited but not in a way that we don't update the limit that was set a decade ago.
We don't use that limit. The upper limit is now 4 MB on disk, although in reality most blocks are around 2 MB.
I think that flexible block size, i.e. block size increases and decreases according to number of transactions
Monero uses dynamic block sizes, and so fees remain low at all times. However, Monero has a tail emission of 0.6 XMR per block for ever more.
Economically speaking, both achieve the same thing - users pay miners. The difference is whether the user making the transaction pays the miner directly (Bitcoin) or whether every user of Monero pays the miner via a very small amount of inflation (Monero). Either way, you need some way to fund miners or the chain becomes insecure.
And I strongly disagree with those who say congestion should exist to help miners due to block subsidy going down. Because even after the upcoming halving the miners would still be paid about $135k per block they find at the current price. This is enough incentive so that they don't need to rely on fees.
For 4 years. Then it halves again. And again. People have this distant figure of "2140" in their head as when the subsidy goes to zero and we can just kick the can down the road and we don't really need to care about fees until then, but in reality most people using bitcoin today are going to live to see the point when the subsidy becomes negligible. It's only going to take 20 years for the block subsidy to fall below 0.1 BTC. Even if we think bitcoin will be $100,000, then you are down to only $10k per block which is around 2-3% of what miners are earning now. And it only goes down from there. Unless you believe bitcoin is going to be worth $10 million or more within in the next 20-30 years, we need a competitive fee market.
I do think that block size will increase again, probably multiple times, but we can't just jump to the initial suggestion of having blocks large enough to let every transaction process at 1 sat/vbyte without greatly decreasing the future security of the network.